Can embodied intelligent robots become "mining machines"? How PrismaX builds a robot coordination layer?
TechFlow Selected TechFlow Selected
Can embodied intelligent robots become "mining machines"? How PrismaX builds a robot coordination layer?
Three-minute read on PrismaX, the embodied AI robot orchestration layer backed by a16z.
Navigating Web3 tides with focused insightsBy KarenZ, Foresight News
In recent years, humanoid robot hardware technology has made significant progress, from dexterous robotic hands to high-precision actuators, with some advanced components achieving commercialization. However, large-scale applications still face critical bottlenecks: software not yet production-ready, data scarcity, high management costs, and low human-robot collaboration efficiency. Most current robotics companies rely on self-built data collection systems, leading the industry into a "data silo"困境 (predicament), limiting the leap of robotic intelligence toward mainstream adoption.
Against this backdrop, PrismaX emerged, aiming to build a decentralized embodied AI coordination layer by connecting stakeholders through open protocols to create an efficient, transparent, and scalable open robotics coordination economy. Recently, PrismaX completed an $11 million funding round led by a16z crypto CSX, drawing significant attention from robotics enthusiasts. So what makes PrismaX so compelling? What enables it to stand out in a fiercely competitive market?
PrismaX Team Background and Investor Lineup
PrismaX was co-founded by Bayley Wang and Chyna Qu, with team members possessing deep expertise and practical experience in robotics technology and decentralized economies.
Bayley Wang, PrismaX Co-Founder and CEO, holds an academic background from the Massachusetts Institute of Technology (MIT) and extensive entrepreneurial experience, focusing on augmented reality, consumer electronics, and robotics. His career demonstrates a successful transition from academic research to commercialization, particularly in technology and hardware development.
-
From 2011 to 2012, Bayley Wang served as a researcher at MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL), developing high-performance optical ray tracing simulation and optimization tools using C/C++. Later, he conducted research at MIT in robotics, autonomous driving, algorithm development, and imaging system design. Bayley Wang also served as an instructor for MIT's Educational Support Program (ESP).
-
Ranked among the top 25 nationally in the United States Mathematical Olympiad.
-
In 2012, while an undergraduate at MIT (sophomore year), founded his first consumer electronics education startup, One Tesla, which achieved over one million dollars in revenue that year and was later acquired.
-
Co-founded Kura Technologies, an AI wearable devices company, from 2019 to 2024, focusing on AR glasses and platform development.
-
Bayerly Wang holds several patents related to embodied intelligent robots and is a co-inventor of "AR Headsets with Improved Micro Structure Display" and "Augmented Reality Eyepiece the Manufacturing Methods."
On the funding front, in mid-June 2025, PrismaX closed an $11 million financing round led by a16z, which alone invested $7 million. Other investors include Stanford Blockchain Accelerator, Symbolic, Volt Capital, and Virtuals Protocol. Notably, PrismaX is part of a16z crypto’s CSX 04 cohort for early-stage crypto startups and officially launched at the a16z CSX Demo Day on June 3.
What Is PrismaX? Core Insights from the Whitepaper
According to the PrismaX whitepaper, PrismaX aims to build an open robotics coordination economy through a decentralized data incentive mechanism and a unified teleoperation (teleop) standard.
In simple terms, PrismaX serves as the “public data and labor layer for the robotics world,” combining teleop protocols, a data engine, a three-party marketplace, evaluation models, and a token-based incentive economy. This allows anyone to remotely operate robots, contribute data, earn token rewards, and continuously generate high-quality training data for AI companies.
PrismaX’s solution is built around three pillars, creating a self-reinforcing “flywheel effect”:
1. Open-source teleop protocol: Connects global teleoperators with robots, enabling operators to control robots via standardized interfaces to complete tasks and generate high-value data.
2. Distributed Data Engine: Accumulated data from the teleop protocol can be used to train AI models. The PrismaX data market divides data ownership into network-shared data and customer-private data:
-
Network-shared data: Controlled by the community. New tokens are minted based on scores from the Eval Engine, forming the core of PrismaX’s innovative “Proof-of-View” mechanism. When data is accessed, part of the transaction fee is burned, and part is redistributed to the data creators.
-
Customer-private data: Collected on-demand and paid per use; post-transaction tokens are redistributed to data creators based on data volume.
Additionally, ownership of ultra-large-scale visual task data belongs to the network. Minting new tokens occurs based on Eval Engine scores when visual data is collected, and accessing datasets burns tokens paid by data requesters.
Notably, PrismaX uses an automated evaluation engine—Eval Engine—to collect visual data and score the quality of robotic operation and visual data within the network. This addresses data credibility issues, incentivizes high-quality contributions, supports data filtering, and helps AI companies quickly identify datasets suitable for training. Specifically, the Eval Engine employs open-source AI models to extract key features—for example, computing CLIP-L and DINOv2 embeddings for each video frame—and considers prediction error detection and effective action identification via optical flow analysis. Scoring dimensions include motion, semantics, aesthetics, and diversity.
3. Three-party marketplace for teleoperators, data buyers, and robot owners: Supports use cases such as data collection, robot leasing, and rentals, ultimately enabling inter-robot coordination and transactions.
PrismaX Economic System
The core design of the PrismaX platform economy aims to solve the robotics industry’s cold-start problem (lack of economic incentives → insufficient robot deployment → limited data → constrained AI model training → low robot utility).
Centered on value creation, distribution, and circulation within the ecosystem, PrismaX uses the PIX token as its core instrument, integrating staking, incentives, token minting, and burning mechanisms. Both minting and burning are tied to real contributions and demand, providing coordinated incentives for teleoperators, robot owners, data contributors, and other participants to drive self-sustaining ecosystem growth.
Teleoperators earn token rewards upon task completion, with faster completion yielding higher reward multipliers. Staking tokens improves reputation and grants priority access to high-yield tasks. In the data market, when network-owned data is accessed or consumed, part of the tokens paid by enterprises (demand side) are burned, while the rest are redistributed to data creators. For transactions involving customer-private data, redistribution to data creators is based on data volume.
Robots on PrismaX function like “mining rigs,” offering owners multiple income streams and transforming the economic model of robot ownership. For instance, robot owners can partner with data clients to provide custom datasets and collect transaction fees.
How to Interact?
PrismaX has launched a points system and robot reservation system, allowing users to earn points through the following steps:
1. Log in using a wallet or email to receive 1,000 initial points plus 10 daily points.
2. Read the whitepaper and complete the quiz to earn 3,500 Prisma points.
3. Daily login grants 10 points.
4. Reserve a robot to receive a 3x points boost ($99, pay-as-you-go).
PrismaX previously noted that users will soon be able to control robotic arms via the PrismaX Gateway, earning points by playing teleop games and completing various tasks.
Summary
PrismaX will focus in its first phase on feeding teleop and visual data into model training. In the second phase, operators will begin taking commercial orders, and robots will enter real production lines. In the third phase, robots will achieve high autonomy, and the PrismaX network will shift toward delivering production-grade services for millions of robots.
As PrismaX CEO Bayley Wang stated, “The PrismaX platform will allow humans to work alongside AI rather than be replaced by it.” PrismaX’s vision is to build a “flywheel effect” through three pillars—data, teleop, and models: large-scale visual data builds better foundation models, enhances teleop efficiency, and drives further real-world data collection, forming a sustainable robotics development ecosystem.
Join TechFlow official community to stay tuned
Telegram:https://t.me/TechFlowDaily
X (Twitter):https://x.com/TechFlowPost
X (Twitter) EN:https://x.com/BlockFlow_News
Foresight News
